CA2566617C - Method and composition for reducing e.coli disease and enhancing performance using bacillus - Google Patents

Abstract

Bacillus strains that inhibit pathogenic swine E. coli and /or improve performance are provided. Inhibition of pathogenic swine E. coli decreases E.
coli disease. At least one strain enhanced swine performance by improving average daily gain, feed efficiency, and feed intake. Preferred Bacillus strains are of species that are included on the GRAS list. Bacillus species are sporeformers and therefore are highly stable and can be fed to swine.

Description

METHOD AND COMPOSITION FOR REDUCING E.COLI
DISEASE AND ENHANCING PERFORMANCE USING BACILLUS
BIBLIOGRAPHY
Complete bibliographic citations of the references referred to herein by a reference numeral in parentheses can be found in the Bibliography section, immediately following the Examples.
FIELD OF THE INVENTION
The invention relates to bacterial strains and the use thereof to control disease in animals and enhance animal performance. More particularly, the invention relates to Bacillus strains and their use to control diseases caused by Escherichia coli and to enhance animal performance.
BACKGROUND OF THE INVENTION
E. coli disease is an important and devastating disease to swine producers.
Known for causing edema disease (ED) and post weaning diarrhea (PWD), the economic impact can be substantial with death losses as high as 50% (1). Multiple management and environmental factors have been associated with E. coli infections, including age at weaning, diet, crowding, and transportation. Certain host genetic factors, such as having receptors for the E. coli fimbriae to attach to the intestinal surface, also contribute to a pig's susceptibility to E. coli infection.
ED and PWD are both caused primarily by hemolytic E. coli proliferating in the small intestine. E. coli infection can also be diagnosed in pigs shortly after birth to two weeks of age (3). ED can occur in pigs between three and eight weeks of age and is characterized by subcutaneous and subserosal edema, a progressive ataxia, paralysis and a high mortality (2).
PWD is commonly observed at 7-10 days post weaning but can occur up to eight weeks of age and is characterized by reduced growth rate, severe diarrhea, dehydration, toxemia, or death (2, 3).
Both PWD and ED can occur in the same group of pigs and the causative E. coli strains often share certain virulence factors. Usually the first manifestation seen with PWD is sudden death, as early as two days, after weaning. Pigs that do not die suddenly, display a decrease of feed consumption and watery diarrhea which leads to depression and life threatening dehydration. Many pigs show cyanotic discoloration of the tip of the nose, the ears, and the abdomen. Staggering and uncoordinated movements may also be seen in severely affected pigs.
Peak mortality generally occurs 6-10 days after weaning. In a herd with PWD, morbidity can 0 vary. Within a litter, the morbidity may be high and reach up to 80% with an average of 30-40%. Mortality in untreated herds can reach 26% (4).
Anorexia is often the first sign seen with edema disease. If diarrhea is to occur, it usually follows after the anorexia. The diarrhea usually disappears by the time the edema and nervous involvement become apparent. Edema can be seen in the eyelids, forehead, ears, and lips. Upon 5 necropsy edema can be seen in the submucosa of the stomach, the mesocolon, gallbladder, and lungs. Progressive ataxia and mental confusion leading to complete recumbence and severe dyspnea are seen in the final stages. The mortality rate in edema disease can reach 50% to over 90% (4).
The source of E. coli in a weanling pig is usually derived from the environment either in tO the nursery, or the pig may acquire the E. coli in the farrowing unit and carry it into the nursery.
Pathogenic E. coli can spread by means of aerosol, feed, farm vehicles, pigs, and other animals (4).
E. coli are part of the normal intestinal flora in pigs. Most of the intestinal E. coli do not possess the ability to cause disease. These E. coli pass through the intestines and are not able to attach to the intestinal wall and do not produce toxins. Those E. coli that are pathogenic and cause disease have the ability to do so because they have obtained genes which code for specific virulence factors (5). These virulence factors allow the E. coli to adhere to the intestinal wall, colonize the intestine, and produce enterotoxins, which can cause diarrhea and verotoxins which can cause edema disease.
Enterotoxigenic E. coli (ETEC) is the major type of E. coli implicated in diarrheal disease of pigs (FIG. 1) (5). These strains are characterized by their ability to adhere to the pig intestine and produce enterotoxins. Adherence to the intestinal tract is performed by frimbriae (pili) on the bacteria that attach to receptors located on the intestinal surface. These pili are highly antigenic filamentous protein structures that extend from the surface of the bacteria (5).
The major pili found on ETEC are F4 (K88), F5 (K99), F6 (987p), F41, and F18.
Enterohemorrhagic E. coil (EHEC) is the major type of E. coil implicated in edema disease of pigs. The basis of colonization and toxin production is the same as with the ETEC.
The only pilus that has been associated with edema disease in swine is F18, with the Fl8ab variant being associated more commonly with edema disease. The toxin produced by the EHEC
is known as Stx2e. This toxin belongs to a family of toxins called shiga toxins or verotoxins. It 0 is a high molecular weight protein that binds to specific receptors on vascular endothelial cells in certain target tissues (5). Therefore, the disease seen with EHEC is a result of toxemia. The receptor for the toxin is found in blood vessels in the brain, eyelid, stomach wall, mesentery of the colon, and the spinal cord (5). The toxin causes injury and death to the endothelial cells in these target organs.
5 Enteropathogenic E. coil (EPEC) strains, also known as attaching and effacing E. coli (AEEC) may play a role in diarrheal disease of pigs. These strains have only recently been investigated as a cause of diarrhea in weaned pigs and were first associated with diarrhea in humans (5). The EPEC cause disease by forming an attachment to the pig intestinal epithelial cells, possibly through the use of pili, and cause destruction of the microvilli (5). The lesions !O are called attaching and effacing lesions.
No universally effective prophylaxis is available for post weaning E. coil disease (4).
Fundamental to the prevention of disease is to prevent villous atrophy and colonization (7).
Villous atrophy and colonization are related to many factors such as rotavirus infection, diet, STb, and stress (7). Managerial factors that contribute to stress are changes in temperature, overcrowding, feed changes, humidity, and mixing of pigs.
No vaccines are currently commercially available for post weaning E. coil disease.
However, there are companies that will prepare for each farm a killed or modified live vaccine using one attenuated strain of pathogenic E. coil found on the farm. This vaccine generally contains F18 or K88 pili but lacks the toxin genes. The attenuated strain is often grown on the ,0 farm and fed or given intranasal to pigs. Toxoids made from Stx2e are also used, but again are not commercially available. These vaccines are often not very pure and even though they may impact mortality due to E. coil disease, they generally do not decrease mortality to acceptable levels.
Egg immunoglobin, produced by hens that were vaccinated against fimbrial E.
coli antigens, have also been developed as an antibody-containing egg powder in pig feed (4). The egg immunoglobin is produced by vaccinating the hen with an attenuated strain of ETEC or with fimbriae from the pathogenic E. coil strains. The egg yolks are then collected, and egg yolk antibody powder is then obtained by freeze drying the water soluble protein fraction of egg yolks (9). The theory was that it would provide immune protection against colonization with K88 and F18 positive E. coil (4). Marquardt et al. showed that egg-yolk antibodies were able to prevent 0 experimentally induced ETEC diarrhea in 3 day old and 21 day old weaned pigs, and also decreased the occurrence of diarrhea in early-weaned pigs during a field trial study (9).
Nevertheless, this is not always what is seen in the field, and producers seem to get mixed results using egg-yolk antibodies. What has been shown is that protection is only provided against challenge strains that have only the F18 fimbriae in common with the vaccine strains (4).
5 Protection may also occur only for the strain of E. coli the hen was vaccinated for. Another drawback is that egg immunoglobin can be expensive to include in pig diets.
Some pigs are genetically resistant to ETEC and EHEC strains because they genetically lack the K88 and/or F18 pili receptor. Breeding for genetic resistance can help control E. coli disease. The difficult part about this process is the expense of testing and lack of tests available.
!O A proprietary test for the presence of F18 receptors has been developed, but no test exists for the K88 receptor (6).
Antibiotics are often added to feed as a preventative. There are many drawbacks such as consumer acceptance and selection of resistant bacteria (4). Numerous antimicrobial substances are used for this purpose; some include: sulfonamide, trimethoprim, gentamicin, and other aminoglycosides. Isolates from ETEC and EHEC show the highest rate of resistance within swine E. coli, and this resistance is often induced within days or a few weeks (4).
Zinc oxide and spray dried porcine plasma included in weanling pig diets have also been used with mixed success.
Once an E. coil outbreak occurs, treatment must be administered to decrease mortality ;0 and morbidity. Antimicrobial therapy has been the treatment of choice.
Antibiotics can be given parenterally or in the water once disease is detected. Antibiotic resistance with E. coli isolates is widely known. Pathogenic E. coil resistance has been detected against every antibiotic approved for use (4). Electrolytes can also be offered as a treatment choice but can be very costly to the producer.
One important factor that could result in the failure of current treatment and prevention techniques is the high genetic diversity of ETEC and EHEC strains. A high degree of heterogeneity has been shown among isolates from the same state and farm (8).
Wilson, et al showed that serotypes associated with post-weaning diarrhea appear to be limited but have very diverse genetic backgrounds (10). This leads one to believe that multiple strains of the same virulence factors or serotype are the cause for a single outbreak of E. coil disease on a farm.
The cause of heterogeneity is uncertain, but may be due to the fact that gene transfer can readily occur within swine E. coil. The fact that antibiotics, vaccines, and other treatments are always being used instigates the need of gene transfer for the survival of pathogenic swine E.
coll. In addition, the great amount of fecal oral transmission in swine systems provides an environment needed for gene transfer to occur. This heterogeneity is evidence to explain why many traditional methods of treatment and prevention fail.
Therefore, what is needed is one or more isolated Bacillus microorganism that is capable of at least one of (A) inhibiting E. coil disease and (B) improving performance of an animal. A method of feeding swine one or more of the above-referenced Bacillus microorganisms to inhibit E. coil disease and/or improve performance of the swine is also needed. Additionally needed is a method of forming a direct-fed microbial from the above-referenced Bacillus microorganisms.
SUMMARY OF THE INVENTION
The invention, which is defined by the claims set out at the end of this disclosure, is intended to solve at least some of the problems noted above. Provided is an isolated microorganism of the genus Bacillus selected from the group consisting of strains 3A-P4 ATCC PTA-6506, 15A-P4 ATCC PTA-6507, and 22C-P1 ATCC PTA-6508. Provided is an isolated microorganism of the genus Bacillus that is capable of at least one of the following: (A) inhibiting E. coil disease and (B) improving performance of an animal. In one embodiment, the microorganism is selected from the group consisting of strains 3A-P4, 15A-P4, and 22C-Pl. In another embodiment, the microorganism is strain 15A-P4. A combination of microorganisms comprising at least two of the above-listed microorganisms is also provided.

Additionally provided is a method of feeding swine. In the method, at least one of the above-listed strains is fed to the swine. There is also provided the use of at least one strain of microorganism of the genus Bacillus, in the feeding of swine, the at least one strain being selected from the group consisting of strains 3A-P4 ATCC PTA-6508, 15A-P4 ATCC
PTA-6507, and 22C-P1 ATCC PTA-6508. A method of forming a direct-fed microbial including at least one of the above-listed strains is also provided. In the method at least one of the above-listed strains is grown in a liquid nutrient broth. The microorganism is separated from the liquid to form the direct-fed microbial. Thus, there is provided a method of forming a direct-fed microbial, the method comprising: (a) growing, in a liquid nutrient broth, at least one microorganism of the genus Bacillus; the at least one strain being selected from the group consisting of strains 3A-P4 ATCC PTA-6508, 15A-P4 ATCC PTA-6507, and 22C-P1 ATCC
PTA-6508; and (b) separating the microorganism from the liquid to form the direct-fed microbial.
BRIEF DESCRIPTION OF THE DRAWINGS
Preferred exemplary embodiments of the invention are illustrated in the accompanying drawings in which:
FIG. 1 is a diagram showing colonization and disease due to Enterotoxigenic E.
coli (ETEC), Enteropathogenic E. coli (EPEC), and Enterohemorrhagic E. coli (EHEC).
FIG. 2 is a dendogram of the E. coli isolates obtained from the diagnostic laboratory combining the RAPD analysis using primer 1 and primer 2. Also shown is the multiplex results reported from a diagnostic laboratory and the multiplex results obtained from Agtech Products laboratory.
FIG. 3 is a gel image of two sets of DNA fingerprints for three preferred strains of Bacillus isolates: 3A-P4 (lanes 1 and 5), 15A-P4 (lanes 2 and 6), and 22C-P1 (lanes 3 and 7). A
100 bp molecular weight marker (Bio-Rad, Hercules, CA) is in lane 4. Two different 10 bp primers were used for two sets of random amplified polymorphic DNA analysis on the isolates, with results of the first set shown in lanes 1-3, and results of the second set shown in lanes 5-7.
FIG. 4 is a graph showing percent inhibition of E. coli strain E.20 by Bacillus isolate 3A-P4 at different optical density (OD) readings.
FIG. 5 is a graph showing percent inhibition of E. coli strain E.20 by Bacillus isolate 15A-P4 at different OD readings.

FIG. 6 is a graph showing percent inhibition of E. coli strain E.20 by Bacillus isolate 22C-P1 at different OD readings.
FIG. 7 is a graph showing percent inhibition of E. coli strains E.20 and E.23 by Bacillus isolate 3A-P4 at different time points.
FIG. 8 is a graph showing percent inhibition of E. coli strains E.20 and E.23 by Bacillus isolate 15A-P4 at different time points.
FIG. 9 is a graph showing percent inhibition of E. coli strains E.20 and E.23 by Bacillus isolate 22C-P1 at different time points.
FIG. 10 is a graph showing growth curves, in the absence of any Bacillus isolate, of E. coli strains E.20 and E.23 at different time points.
FIG. 11 is a SDS-PAGE on 15A-P4 after ammonium sulfate precipitation. P=
pellet fraction, S= supernatant fraction. The box outlines the suspected inhibitory protein located between the 31,000 and 45,000 molecular weight marker. The (+) denotes inhibition of E. coli using the spot plate method. The (-) denotes no inhibition of E. coli using the spot plate method.
FIGS. 12A and 12B are the graphs showing the mode of action of the active metabolite produced by 3A-P4 on E. coli strain E.20 (FIG. 12A) and E. coli strain E.23 (FIG.
12B).
FIGS. 13A and 13B are the graphs showing the mode of action of the active metabolite produced by 15A-P4 on E. coli strain E.20 (FIG. 13A) and E. coli strain E.23 (FIG. 13B).
FIGS. 14A and 14B are graphs showing the mode of action of the active metabolite produced by 22C-P1 on E. coli strain E.20 (FIG. 14A) and E. coli strain E.23 (FIG. 14B).
FIGS. 15A and 15B are graphs showing the effect of a preferred embodiment of a combination of Bacillus strains (Product 3) on pig weight for Field Trial Eat Day 7 (FIG. 15A) and Day 15 (FIG. 15B).
FIG. 16 is a graph showing the effect of Product 3 on ADG for Field Trial E.
FIG. 17 is a graph showing the effect of Product 3 on feed intake for Field Trial E.
FIGS. 18A and 18B are graphs showing the effect of Product 3 on mortality for Field Trial E at Days 0-7 (FIG. 18A) and Days 0-28 (FIG. 18B).

FIG. 19 is the multiplex results of Field Trial E.
Before explaining embodiments of the invention in detail, it is to be understood that the invention is not limited in its application to the details of construction and the arrangement of the components set forth in the following description or illustrated in the drawings. The invention is capable of other embodiments or being practiced or carried out in various ways.
Also, it is to be understood that the phraseology and terminology employed herein is for the purpose of description and should not be regarded as limiting.
7a DETAILED DESCRIPTION
Definitions:
The following definitions are intended to assist in providing a clear and consistent understanding of the scope and detail of the terms:
As used herein, "active metabolite" means a substance produced by bacteria and which has antibacterial activity towards other genuses of bacteria.
As used herein, "animal" means a multicellular organism of the kingdom Animalia.
As used herein, "bacteriocin" means a substance produced by bacteria and which has antibacterial activity towards other genuses of bacteria.
0 As used herein "basemix" or "concentrated basemix" refers to Bacillus strains added to a carrier to make a basemix form. The concentrated form is composed of the Bacillus strains added the carrier in a more concentrated form. The basemix or concentrated basemix forms are then be added to the feed at a desired inclusion rate and fed to the animal.
As used herein, "performance" refers to the growth of an animal, such as a pig, measured [5 by the following parameters: average daily gain (ADG), weight, mortality, feed conversion, which includes both feed:gain and gain:feed, and feed intake.
"An improvement in performance" as used herein, means an improvement in at least one of the parameters listed above under the performance definition.
In accordance with the present invention there may be employed conventional molecular )..0 biology and microbiology within the skill of the art. Such techniques are explained fully in the literature. See, e.g., Sambrook, Fritsch & Maniatis, Molecular Cloning: A
Laboratory Manual, Third Edition (2001) Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
Aerobic and facultative sporeformers of the genus Bacillus were isolated.
Bacillus species are the only sporeformers that are considered GRAS, i.e., generally recognized as safe.
Z5 In a preferred embodiment, a Bacillus species was included only if it had GRAS status. The Bacillus species were isolated from environmental samples such as poultry litter and animal waste and screened. Other sources of Bacillus can also be screened. The Bacillus strains were screened for their ability to inhibit growth of pathogenic swine E. coli.
Although not intended to be a limitation to the present disclosure, it is believed that inhibition is accomplished via the 30 secretion of an active metabolite from the Bacillus. While applicants do not wish to be restricted to a particular theory of how the active metabolite inhibits microbial growth and do not intend to limit the present disclosure, it is believed that the active metabolite is a proteinaceous substance and, more specifically, it is believed to be a bactericidal.
The Bacillus isolates were tested for their ability to inhibit various pathogenic strains of swine E. coll. The E. coli strains were obtained from animal diagnostic laboratories, swine environment, and fecal matter. The E. coli strains were shown to be pathogenic by performing multiplex PCR to detect pili and toxin genes associated with pathogenic E.
coli disease in swine.
To test for production of active metabolite, Bacillus isolates were replica plated onto pathogen indicator plates, which were formed from a 1% pathogen inoculum of a pathogenic swine E. coli strain.
Additionally, the active metabolite activity of the Bacillus isolates was reconfirmed using a spot assay method. The Bacillus isolates were then tested using appropriate biochemical tests to determine whether the isolates had GRAS status.
The spectrum of activity of the various Bacillus isolates was then determined.
The Bacillus isolates were tested for activity against the known swine E. coli pathogens collected from different regions of the United States using the spot assay method. This was done to confirm that the activity produced by the Bacillus isolates would be useful throughout the United States. The Bacillus isolates that showed the highest inhibitory activity against numerous pathogenic E. coli strains were further characterized for their activity level.
From these experiments, three preferred Bacillus strains were found: 3A-P4, 15A-P4, and 22C-P1, although other strains can also be used. These strains were preferred because of the number of pathogenic E. coli strains that each inhibited and because of their GRAS status. On January 12, 2005, strains 3A-P4, 15A-P4, and 22C-P1 were deposited at the American Type Culture Collection, 10801 University Blvd., Manassas, VA 20110-2209 and given accession numbers PTA-6506 (3A-P4), PTA-6507 (15A-P4), and PTA-6508 (22C-P1). The deposits were made under the provisions of the Budapest Treaty on the International Recognition of the Deposit of Microorganisms for the Purposes of Patent Procedure.
Strains 3A-P4, 15A-P4, and 22C-P1, can be fed individually or in combination, to swine, although other strains are included within the scope of the invention. For the Bacillus strains 3A-P4, 15A-P4, and 22C-P1, growth times were determined for production of an optimal level of the active metabolite using the broth activity method. The assay incubation time at which optimal inhibition of E. coli occurred was also determined using the broth activity method. For this, active metabolite was added to a culture of E. coli and ODs were read at various time points. The Bacillus isolates 3A-P4, 15A-P4, and 22C-P1 were tested for activity against 142 known swine E. coli pathogens collected from different regions of the United States using the broth activity method. These strains inhibited to varying degrees one hundred percent of the E.
coli strains tested. The strains can be used alone or in combination to inhibit growth of pathogenic swine E. coll.
The three preferred Bacillus isolates of the invention were isolated from different geographical regions of North America and from different environmental sources. Specifically, strain 3A-P4 was isolated from chicken litter from Canada, strain 15A-P4 was isolated from turkey litter from the Western United States, and strain 22C-P1 was isolated from a swine lagoon from the Eastern United States.
The active metabolite was then purified from each of the Bacillus isolates to two levels:
first, a crude purification of the active metabolite was obtained by filtering the culture supernatant, and second, a partially purified active metabolite was obtained by salting out the l 5 active metabolite and then fractionating it by column chromatography.
The stability and characterization of the active metabolite was then determined using the crude form of the active metabolite by performing enzyme and heat degradation assays, mode of action assays, and antibiotic sensitivity assays. Optimal media and time conditions were also determined for cell growth and spore formation.
Further characterization of the Bacillus isolates 3A-P4, 15A-P4, and 22C-P1 was performed, including DNA fingerprinting and determining stability of the isolates in a swine premix at 60 C for 8 weeks.
Through field trial studies, it was determined that the Bacillus strains also enhance nursery swine performance. Therefore, it is economical for a producer to routinely include the Bacillus strains, either individually or in combination, in feed not only to prevent E. coli disease but also to enhance performance.
The Bacillus isolates of the invention, which inhibit pathogenic E. coli, can be directly fed to swine to inhibit pathogenic swine E. coli disease and to enhance swine performance.
Feeding microorganisms that have GRAS status to livestock is an acceptable practice amongst producers, veterinarians, and others in the livestock industry. By inhibiting this pathogen in the swine, the Bacillus isolates reduces and even prevents E. coli disease in swine.

The Bacillus isolates can be administered as a preventative to swineherds not currently infected with pathogenic E. coli. In a preferred embodiment, newly weaned pigs are fed the Bacillus isolates throughout the nursery phase to inhibit or even prevent outbreaks of E. coli disease and to enhance performance. However, one or more Bacillus isolate can be fed at other phases also. Routine administration of the microorganisms dramatically reduces and even eliminates outbreaks of E. coli disease at animal production facilities and enhances swine performance.
The Bacillus isolates can be administered as a direct-fed microbial.
Administration of one or more direct-fed microorganisms to animals is accomplished by any convenient method, [0 including adding the Bacillus isolates to the animals' drinking water or to their feed, or by direct oral insertion. In a preferred embodiment, the microorganism is fed to animals by adding it to the animals' feed or water. Bacillus isolates preferably are administered as spores.
The Bacillus isolates may be presented in various physical forms, for example as a top dress, liquid drench, gelatin capsule, gel, or added to the water. In feed form, the isolates may be presented in a form as a basemix or a concentrated form of the basemix. In a preferred embodiment of a top dress form, freeze-dried Bacillus fermentation product is added to a carrier, such as whey, maltodextrin, sucrose, dextrose, limestone (CaCO3), rice hulls, yeast culture, dried starch, or sodium silico aluminate.
In a preferred embodiment of the liquid drench form, freeze-dried Bacillus fermentation ZO product is added to a carrier, such as whey, maltodextrin, sucrose, dextrose, dried starch, or sodium silico aluminate, and a liquid is added to form the drench.
In a preferred embodiment of the gelatin capsule form, freeze-dried Bacillus fermentation product is added to a carrier, such as whey, maltodextrin, sugar, limestone (CaCO3), rice hulls, yeast culture dried starch, or sodium silico aluminate. The Bacillus isolates and carrier can be .Z5 enclosed in a gastrointestinal-degradable gelatin capsule.
In a preferred embodiment of the gel form, freeze-dried Bacillus fermentation product is added to a carrier, such as vegetable oil, sucrose, silicon dioxide, polysorbate 80, propylene glycol, butylated hydroxyanisole, and citric acid, or artificial coloring to form the gel. In a preferred embodiment of the water form, the freeze-dried Bacillus fermentation product is added 30 to a carrier, such as sucrose, dextrose, sodium silico aluminate, and artificial coloring.

In a preferred embodiment of the basemix or concentrated basemix form, the freeze-dried Bacillus fermentation product is added to a carrier, such as, but not limited to, rice hulls, dried brewers grain, limestone, or baylith for moisture control. Other carrier that are suitable and compatible for these strains can also be used.
The microorganisms can be administered as spray-dried, freeze-dried, fluidized bed dried, used in a solid state fermentation form, as well as other forms. For freeze drying, the wet cell paste preferably is then mixed with cryoprotectants, which maintain the viability of the cells during the freezing and drying process. The mixture is then placed in trays, frozen and subsequently dried. For spray drying, the paste or slurry is then mixed with spray drying aids, if [0 applicable, and spray dried. The resulting dried cake, obtained from drying methods or solid state fermentation method, is then milled to a uniform size and plated to determine the activity.
After a viable cell count has been determined, the cell count preferably is standardized to a predetermined activity level or colony forming units (CFU) per gram by blending with dry carriers.
To produce the Bacillus isolates, one or more isolates are grown in a liquid nutrient broth, such as TSB, preferably to a level at which the highest number of spores are formed. In a preferred embodiment, the isolates are grown to an OD where the spore yield is at least 1 x 109 colony forming units (CFU) per ml of culture. The bacterial strains of the present invention are produced by fermentation of the bacterial strains, fermentation is started by scaling-up a seed 0 culture. This involves repeatedly and aseptically transferring the culture to a larger and larger volume to serve as the inoculum for the fermentation, which is carried out in large stainless steel fermentors in medium containing proteins, carbohydrates, and minerals necessary for optimal growth. A non-limiting exemplary medium is TSB. After the inoculum is added to the fermentation vessel, the temperature and agitation are controlled to allow maximum growth.
).5 Once the culture reaches a maximum population density, the culture is harvested by separating the cells from the fermentation medium. This is commonly done by centrifugation. The count of the culture can then be determined.
The count of the bacteria is important when combined with a carrier. At the time of manufacture of the composition, the Bacillus count preferably is at least about 1.0 x 1011 CFU/g.
;0 The counts may be increased or decreased from these base numbers and still have complete efficacy. CFU or colony forming unit is the viable cell count of a sample resulting from standard microbiological plating methods. The term is derived from the fact that a single cell when plated on appropriate medium will grow and become a viable colony in the agar medium.
Since multiple cells may give rise to one visible colony, the term colony forming unit is a more useful unit measurement than cell number.
To prepare the compositions, the cultures and the carrier can be added to a ribbon or paddle mixer and mixed preferably for about 15 minutes. The components are blended such that a uniform mixture of the carrier and cultures result. The final product is preferably a dry flowable powder. Exemplary carriers in this composition are rice hulls, dried brewers grain, limestone, baylith, or other suitable carriers for microorganisms.
0 The preferred dosage range of the liquid drench, gelatin capsule, and gel is about 1x104 CFU/g or ml/day to about lx101 CFU/g or ml/day, and more preferably about lx106 CFU/g or ml/day. The preferred dosage range of the top dress, basemix, and premix is about lx103CFU/g of feed to about lx108CFU/g of feed, and more preferably about lx106CFU/g of feed. The preferred dosage range for inclusion into water is about lx103CFU/pig/day to about lx101 5 CFU/pig/day, and more preferably about lx108CFU/pig/day. While these examples use freeze-dried Bacillus as an ingredient in the top dress, liquid drench, gelatin capsule, gels, water, and feed forms it is not necessary to freeze-dry the Bacillus before feeding it to swine. For example, spray-dried, fluidized bed dried, or solid state fermentation Bacillus or Bacillus in other states may be used. The microorganisms can also be administered in a wet cell slurry paste, with or !O without preservatives, in concentrated, unconcentrated, or diluted form.
=
The composition used in the Examples below was produced as follows: strain 3A-P4 with a count of 7x10" CFU/g, 15A-P4 with a count of 8.4x1011CFU/g, and 22C-P1 with a count of 6x10" CFU/g were combined in different ratios with carriers to determine the best ratio to inhibit pathogenic swine E. coli and enhance nursery performance. The combinations were as follows: Product I: 30% of the total count of strain 3A-P4, 60% of the total count of strain 15A-P4, 10% of the total count of strain 22C-P1 for a final bacteria count of 5.1x108 CFU/g; Product 2: 100% of the total count of strain 22C-P1 with a final bacteria count of 4.5x108 CFU/g; and Product 3: 90% of the total count of strain 22C-P 1, 10% of the total count of strain 15A-P4 with a final bacteria count of 5.1x108 CFU/g. In the Examples, all the above combinations were 30 added to the feed for a final count of lx106CFU/g of feed and were fed throughout the nursery phase. Carriers used in all combinations used in the experiments were 40% rice hulls, 19% dried brewers grain, 40% limestone, and 1% baylith for moisture control. A preferred combination is 90% of the total count of strain 22C-P1 and 10% of the total count of strain 15A-P4. This combination was found to be effective in decreasing swine E. coli disease and enhancing swine performance.
Additional combinations and single-strain compositions that are useful include an about 90% of the total count of strain 15A-P4 and about 10% of the total count of strain 22C-P1 combination with a final bacteria count of about 1 x 106 CFU/g of feed and 100% 15A-P4 with a final bacteria count of about 1 x 106 CFU/g of feed.
Bacillus strains provided herein are capable of at least one of the following in swine: (A) ' 0 inhibiting E. coli disease and B) improving performance. One or more of the Bacillus strains have been shown to be effective for these purposes when fed to nursery pigs.
It is believed that feeding one or more Bacillus strain provided herein would also be useful when fed to pigs at other stages in their lives. For instance, it is believed that feeding one or more Bacillus strain to breeding stock, including sows, gilts, and boars, and to lactation-phase piglets, and finishing pigs [5 would also provide benefits of inhibiting E. coli disease and/or improving performance.
One or more of the Bacillus strains has also been shown to be beneficial when fed to poultry. For example, populations of avian pathogenic E. coli have been reduced when strain 15-A-P4 was fed to poultry.
!O EXAMPLES
The following Examples are provided for illustrative purposes only. The Examples are included herein solely to aid in a more complete understanding of the presently described invention. The Examples do not limit the scope of the invention described or claimed herein in any fashion.

Isolation of Active Metabolite Producing Bacillus Strains and Isolation of Pathogenic Strains:
A. Bacillus strains and media:
Microbial strains from chicken litter, turkey litter, swine waste, and dairy waste, were screened for Bacillus strains. The environmental sample was weighed and mixed with sterile peptone blanks to make a 10-1 dilution. To pasteurize and, thus, to select for aerobic and facultative sporeformers, the sample was placed in a masticator for one minute and then heated 0 for thirty minutes in a 63 C water bath. The sample was then serially plated onto Tryptic Soy Agar (TSA) and incubated at 32 degrees for 24 ¨ 48 hours to obtain isolates.
B. Pathogen E. coli strains and media:
An initial collection of pathogen strains was obtained from a swine diagnostic laboratory.
5 This included two K88 and one F18 E. coli strains. The isolates were stored as frozen stocks at ¨
85 C in TSB supplemented with 10% glycerol. For the initial screening process, the three E. coli strains were utilized in assays to screen Bacillus isolates for activity against E. coli. Swine E. coli pathogens tested were grown overnight in tryptic soy broth (TSB) at 37 C. A
one percent inoculum was transferred into its corresponding media the next day and incubated at 37 C until ;0 OD value at 600nm was at 0.600.

Activity of Bacillus Isolates Against E. coli:
A. Zone of inhibition assay:
Activity against E. coli was determined by replica plating the Bacillus isolates onto indicator plates containing swine E. coli pathogens. Pathogen indicator plates were formed by transferring one percent of one of the swine E. coli pathogen inoculum, grown as described above, into tempered TSA. Seven milliliters of this agar was poured into a petri dish to make the ,0 pathogen indicator plates. The Bacillus isolates were replica plated onto pathogen indicator plates and were incubated overnight at 32 C. Plates were then observed for zones of inhibition for each pathogen. The Bacillus isolates that produced zones of inhibition were picked off the plate and grown in TSB to isolate the colony for reconfirmation of its activity. The isolates were stored as frozen stocks at ¨85 C in TSB supplemented with 10% glycerol. Thirty thousand Bacillus isolates were screened for activity against E. coil. Fifty isolates produced activity against E. coil.
B. Spot assay:
Activity of the 50 Bacillus isolates was confirmed using the spot plate assay method by growing the isolates in TSB overnight at 32 C. Ten microliters of the Bacillus isolate was then 0 spot plated onto pathogen indicator plates made as described above.
Indicator plates were incubated 24 hours at 32 C and were observed for zones. Thirty six of the fifty isolates displayed activity against E. coli upon reconfirmation using the spot assay method.
Table 2 shows the Bacillus strains that had activity against one or more of the E. coli strains.
It should be noted that some Bacillus strains had activity against more than one E. coil strains.

Table 2. Summary of Active Metabolite Producing Bacillus Isolates Against E.
coll.
INDICATOR STRAIN POSITIVE BACILLUS
STRAINS
Escherichia coli K88 strain BH 15 Escherichia coil K88 strain S 9 Escherichia coli F18 strain 8A133 16 0 Biochemical Tests on Bacillus Isolates:
a All Bacillus isolates that were confirmed to produce active metabolites were biochemical tested to identify isolates that were generally recognized as safe (GRAS).
Testing was performed by both (1) traditional lab methods, including Gram stain, colony morphology, 5 catalase production, starch utilization, casein utilization, nitrate reduction, indole formation, Voges-Proskauer, gelatin hydrolysis, and citrate production and (2) an API
Bacillus biochemical test kit available from bioMerieux of Hazelwood, Missouri. Thirty-six Bacillus isolates were screened using traditional biochemical methods, which showed that six of the thirty-six isolates tested as possible GRAS strains. These isolates were retested using the API
test kit, and results showed that all six isolates were confirmed as belonging to species that are GRAS. The isolates producing the widest spectrum of activity against E. coli were biochemical tested using an outside reference laboratory for final identification.

Determination of Spectrum of Activity Against E. coli of Bacillus Isolates:

A. Spot plate procedure to determine spectrum of Activity Against E. coli:
To confirm that the six GRAS Bacillus isolates were also effective against E.
coli pathogens from other regions of the United States, a broad collection of porcine E. coli pathogens were selected from various animal diagnostic laboratories from Indiana, Oklahoma, 5 Iowa, and Texas. Nineteen other GRAS Bacillus isolates that demonstrated against other pathogens were also screened against diagnostic laboratory E. coli pathogens.
Some of these additional strains are included Table 4 below.
Upon arrival, pathogens were immediately grown in either TSB overnight at 37 C.
Streak plates of each culture were made to ensure pure colonies, which were grown again in their !O respective media overnight at 37 C.
Forty porcine E. coli isolates were obtained. E. coli strains represented were K88, K99, 987p, F18, F41, and 718. Toxins produced by the E. coli included the enterotoxins LT (heat-labile enterotoxin), Sta (heat-stable enterotoxin a), and STb (heat-stable enterotoxin b); and the Shiga-like toxin Stx2e (subgroup Stx2 with variant form e). Thirty-seven of the forty E. coli isolates were confirmed as E. coli using an API biochemical test kit.
The twenty-five GRAS Bacillus isolates, which included the six isolates that showed activity against the original three E. coli isolates and the nineteen isolates that showed activity against other pathogens, were tested for activity against the thirty-eight E.
coli porcine pathogens, using the spot assay method described above in Example 2B. For each Bacillus O isolate, an individual colony was grown overnight in TSB at 32 C and was used as the antimicrobial producer culture.

Pathogen indicators were prepared as follows: E. coli was grown in TSB for 24 hours at 37 C. After 24-48 hours of growth, the pathogen indicators were transferred at 1% into new media and incubated at 37 C until an OD of 0.6-1.0 at 600nm was achieved. The pathogen indicator plates were prepared as described above. Five microliters of the Bacillus culture was spot plated onto the indicator plate and incubated for 24 hours at 32 C. Then, plates were observed for zones of inhibition.
As shown in Table 3 18 of the 25 Bacillus strains displayed inhibitory activity against the thirty-seven E. coli isolates, indicating that an active metabolite that inhibited the pathogen was being produced. Strain 3A-P4 inhibited eleven E. coli isolates, 15A-P4 inhibited fourteen E. coli 0 isolates, and 22C-P1 inhibited eighteen E. coli isolates. Strains 3A-P4, 15A-P4, and 22C-P1 were chosen for further characterization due to their increased activity against E. coli compared to the other 25 GRAS isolates. The three strains 3A-P4, 15A-P4, and 22C-P1 were tested against >140 pathogenic swine E. coli isolates. Each of the three Bacillus strains inhibited the >140 E.
coli isolates by differing percentages.
[5 Table 3. Number of E. coli Strains Inhibited by Bacillus Isolates.
Bacillus Isolate E. coli B. Characterization of the E. coli isolates:
Multiplex, RAPD, and pulse-field gel electrophoresis was performed on the E.
coli isolates obtained from diagnostic laboratories. Upon arrival the isolates were grown and stored as previously described.
All the E. coli isolates were genotyped by Agtech Products using multiplex PCR.
This procedure distinguished E. coli that contained virulence factors responsible for causing disease in swine. Purified genomic DNA was isolated using a DNA
isolation kit (Roche, Indianapolis, IN). The multiplex procedure was performed using the Amplitaq Gold DNA polymerase reagents (Roche, Branchburg, New Jersey) (7). Nine oligionucleatide primers were used in the multiplex procedure to detect the STX2e, LT, STa, and STb toxins; and the K88, K99, F18, 987p, and F41 pili. DNA fragments were separated using a 3.0% Nusieve 3:1 agarose gel (Biowhittaker, Rockland, ME).
The pathogenic E. coli isolates were genetically analyzed using the RAPD
method.
E. coli isolates were grown in TSB overnight until an OD of 4.0 at 600nm was obtained.
Purified genomic DNA was isolated using Roche Molecular Biochemicals DNA
isolation kit (Indianapolis, IN). Once the DNA was isolated, RAPD analysis was performed using Ready-To-Go RAPD Analysis Bead kit from Amersham Pharmacia Biotech (Piscataway, NJ). RAPD analysis was performed using two 10-base pair oligionucleatide primers in a polymerase chain reaction. The DNA fragments were separated using a 1.5%
agarose gel in 0.5X TBE buffer at 100 volts.
Pulsed-field gel electrophoresis (PFGE) was performed using chromosomal DNA embedded in agarose beads and digested with Xba I via a modification of the method of Rehberger (12). DNA fragments were separated on 0.8% agarose gel using a CHEF-DR HI electrophoresis system (Bio-Rad).
DNA bands were visualized following ethidium bromide staining and digitally captured using Syngene Genesnap darkroom software (Frederick, MD). The determination of the molecular size of the DNA fragments and dendogram were accomplished using Bionumerics software (Kortrigjk, Belgium).
The population of E. coli isolated from infected swine herds was heterogeneous, as is shown in the Dendogram of FIG. 2. RAPD analysis differentiated the 48 isolates into 48 genotypic patterns. Of the 48 genotypes, 12 clusters containing 2 or more isolates were identified at a coefficient of similarity of 65% or greater. The largest cluster contained 7 isolates. Eight isolates had less than 67% similarity to any other isolate.
PFGE of intact chromosomal DNA digested with Xba I differentiated 42 isolates into 42 genotypic patterns. Six of the 48 isolates did not produce discernible fragments using PFGE. Of Characterization of Bacillus strains 3A-P4, 15A-P4, and 22C-P1:
A. DNA fingerprinting:
DNA fingerprinting of strains 3A-P4, 15A-P4, and 22C-P1 was performed by random amplified polymorphic DNA (RAPD) PCR analysis on genornic DNA isolated from strains 3A-P4, 15A-P4, and 22C-P1 using Roche Molecular Biochemicals DNA isolation kit (Hoffmann-La B. Stability in premix:
25 Strains 3A-P4, 15A-P4, and 22C-P1 were added at 5.0 x 108 to 500g of swine ration premix. Premix was incubated in a 60 C drying oven for eight weeks. Spore enumeration was performed weekly at 10-6, 10-7, and 10-8.
No decrease in spore count was evident during this assay. Strains 3A-P4, 15A-P4, and 22C-P1 were viable in premix rations containing primarily minerals at temperatures that may be C. Antibiotic sensitivity:
Strains 3A-P4, 15A-P4, and 22C-P1 were grown in 50m1 of TSB overnight at 32 C
with shaking. 1.0% of each strain was inoculated individually into TSA and poured into petri dishes.
After the plates were solidified, the antibiotic discs were placed onto the agar surface. The plates were incubated at 32 C overnight. Inhibitory zones were measured in millimeters.
The antibiotics caused minimal sized zones of inhibition against the three Bacillus strains (Table 4). Therefore, the antibiotics should not interfere with the growth of 3A-P4, 15A-P4, and 22C-P 1 .
Further testing by broth and agar assays was performed with the antibiotic ASP-[0 (chlortetracycline, sulfathiazole, and penicillin combined) against 3A-P4, 15A-P4, and 22C-P 1 .
This antibiotic decreased the growth of 3A-P4, 15A-P4, and 22C-P1 by 99.9%.
This antibiotic is bactericidal to the Bacillus strains.

Table 4. Antibiotic Sensitivity Assay of Strain 3A-P4, 15A-P4, and 22C-P1.
Inhibitory Zones Measured in Millimeters Antibiotic Disc 3A-P4 15A-P4 22C-P1 Oxytetracycxyline 301..tg 24 14 24 Tetracycline 51.tg 26 14 21 Gentamycin 101.1g 26 22 22 Neomycin 51.1.g 16 12 12 Penicillin 2 IU 0 11 13 Bacitracin 10 IU 8 8 8 Lincomycin 21.tg 8 9 8 Biochemical testing of the three Bacillus Isolates, 3A-P4, 15A-P4, and 22C-P1:
Strains 3A-P4, 15A-P4, and 22C-P1 were further biochemical tested using MIDI
laboratory. Ribosomal DNA analysis using Genbank showed that these strains were Bacillus subtilis strains. Bacillus subtilis is a species of Bacillus that is considered GRAS.
?.0 EXAMPLE 7 Purification of Active Metabolite from Bacillus Isolates and Characterization of Activity Against E. coli:
A. Crude purification of active metabolite:
To further optimize the Bacillus strains for optimal active metabolite production and to further characterize the active metabolite for identification a new assay was introduced. This assay, called the broth activity assay, involved using the active metabolite in a crude form so further characterization and optimization tests could be performed without interference from the Bacillus cells. This assay also allows a more quantified result that is reported as percent 30 inhibition. An individual colony of each Bacillus isolate was picked and inoculated into 50 mls of TSB and incubated at 32 C with shaking overnight. After 18 hours of incubation, 10 mls of the producer strain was harvested by centrifugation at 5000 rpm for 10 minutes, and the supernatant was filtered through a 0.2 um acrodisc filter. The filtered supernatant, the crude purified form of the active metabolite, was utilized immediately or stored frozen for no longer than two days before being used in an assay. From an isolated colony, E. coli pathogens were grown in TSB at 37 C, with at least two 1% transfers until an OD of 0.6 at 600nm was reached.
A test tube with TSB was inoculated with the crude form of the active metabolite and the E. coli inoculum. A separate test tube with TSB was inoculated with only the E. coli inoculum and incubated at 37 C. Percent inhibition was determined as follows: (0% OD -sample OD)/ 0%
0 OD*100. To perform this assay correctly, the amount of the crude form of active metabolite needed in the assay, optimal growth time, and OD for the Bacillus isolates to produce an optimal active metabolite level, and the incubation time of the assay was determined.
B. Active metabolite percentage needed for the broth activity assay:
=
5 Trials were performed to determine the optimal percentage of active metabolite needed to inhibit E. coli in the broth assay. Bacillus isolates 3A-P4, 15A-P4, and 22C-P1 were tested against the swine pathogenic E. coli strains. E. coli isolates E.20 and E.23 were chosen to be used as screening pathogens because all three preferred Bacillus isolates (3A-P4, 15A-P4, and 22C-P1) had shown inhibitory activity against these pathogens. From an isolated colony, ,0 pathogens were grown in TSB at 37 C, with at least two 1% transfers until an OD of 0.6 at 600nm was reached. TSB tubes were inoculated at 1% with the pathogen and 10%, 5%, 1%, 0.5%, and 0% with the crude purified form of the active metabolite, collected as previously described, and incubated at 37 C. An OD was read at 4 and 8 hours to determine percent inhibition at each active metabolite percent level. Percent inhibition was determined as follows:
;5 (0% OD - sample OD)/ 0% OD*100. Results are shown in Table 5. The active metabolite added at 10% showed the greatest inhibition of E. coli. Inhibition was also observed at the other levels.
From these results, it was determined the active metabolite added at 10% would yield enough inhibition to be detected with further characterization and optimization tests. Therefore, this percent was used in subsequent studies.

Table 5: Inhibition of E. coli in Broth Using Different Percentages of Active Metabolite Bacillus Inoculated Percent Inhibition Percent Inhibition Isolate Metabolite E. coli E. coli Percentage E. 20 E. 23 4h 8h 4h 8h 3A-P4 10 18.8 26.7 32.6 20.8 9.0 13.3 5.3 8.3 1 0 6.7 5.3 0 0.5 0 6.7 0 0 15A-P4 10 13.6 7.1 15 0 5 9.1 0 10 0 0.5 0 0 5 0 22C-P1 10 13.6 7.4 15.8 4.3 5 9.1 3.7 5.3 4.3 1 0 0 5.3 0 5 C. Active Metabolite Production Time:
Incubation time trials were performed to determine the optimal growth time and OD for the Bacillus isolates to produce an optimal active metabolite level. Percent inhibition against E.
coli strain E.20 was performed and used as the indicator strain. Strains 3A-P4, 15A-P4, and 22C-P1 were used.
0 Growth times and ODs under which the Bacillus isolates were found to produce the most active metabolite and therefore display the greatest amount of inhibition against E. 20 were determined by sampling the culture at 12-hour intervals. At the time points, the crude purified form of the active metabolite was obtained as previously described and inoculated at 10% into 10m1 of TSB. Strain E. 20 was used as the indicator organism and grown as described 5 previously. Strain E.20 was added at 1% to the TSB tubes containing the active metabolite. The assay was incubated at 37 C and OD read at 5h and 10h. A control tube with the indicator organism, added at 1% to a 10m1 TSB tube, was also included in the assay to determine percent inhibition. Percent inhibition was calculated using the formula: (control OD -sample OD)/control OD x 100 for each assay time period, and the average between the 5hr and 10 hr percent inhibition was determined. Results of this are shown in Table 6.
Table 6. The Effect of Incubation Time of Bacillus isolates on E. coli Inhibition Producer 3A-P4 15A-P4 22C-P1 Incubation % Inhibition % Inhibition % Inhibition Time E.20 E.20 E.20 14 h 12.5 7.5 7.5 24 h 26.3 16.3 31.3 36 h 44.0 37.0 51.0 48 h 6.0 9.0 21.0 60 h 13.0 58.4 32.0 72h 1.7 41.3 0.
84 h 0.0 26.1 4.4 For growing of the Bacillus isolates, the OD that produced an optimal level of inhibition against E. coli strain E.20 was also determined. The following ODs were found to produce an 0 optimal active metabolite level: 3A-P4 grown to an OD of 2.5 (FIG. 4), 15A-P4 grown to an OD
of 3.6 to 4.15 (FIG. 5), and 22C-P1 grown to an OD of 2.04 (FIG. 6). At these ODs, the active metabolite produced by the Bacillus isolates produce the most inhibition against E. coli.
Therefore, these growth times should correlate to the highest level of active metabolite being formed.
5 The results obtained determined the time (Table 7) and OD at which the highest percentage of active metabolite was formed for each of the producing strains.
This information was used to grow the producing strains for the remainder of the characterization and optimization tests.
Table 7: Growth of Active Metabolite Producing Bacillus Isolates.
Producer Growing 3A-P4 OD(600nm) 15A-P4 OD(600nm) 22C-P1 OD(600nm) Time in Hours 14 1.62 1.56 1.62 24 2.34 1.56 1.92 362.50 2.04 2.04 =
48 2.22 2.28 2.1 60 2.7 3.6 2.88 72 3.06 4.14 4.2 84 4.2 3.96 4.5 D. Broth activity assay incubation time:
Incubation trials were performed to determine the time of optimal active metabolite inhibition against E. coil using the broth activity assay method. Strains 3A-P4, 15A-P4, and 22C-P1 were used as producer strains, and E.20 and E.23 were used as pathogen indicator strains. The active metabolite was obtained by growing the Bacillus strains to their respective ODs and obtaining the crude purified form of the active metabolite as described previously. The crude purified form of the active metabolite was inoculated at 10% into 10 mls of TSB.
Pathogen indicator strains were grown as previously described and after reaching an OD of .60 (600nm), were added at 1% to the TSB tubes containing the active metabolite.
The assay was incubated at 37 C, and an OD was read at 2, 4, 6, 21, 23, 26, 28, and 30 hours. A control tube with the indicator organism, added at 1% to a 10m1 TSB tube, was also included in the assay to determine percent inhibition and growth pattern. Percent inhibition was calculated as previously described.
Strain 3A-P4 obtained the highest inhibition after four hours (FIG. 7) and, strains 15A-P4 and 22C-P1, after six hours (FIGS. 8 and 9). The E. coil growth curve confirmed that between four and six hours of assay time, the E. coli was at its highest growth (FIG.

10). Thus, inhibition by the active metabolites is not due to the E. coli decreasing naturally.
These results determined the time of highest inhibition of pathogen by the active metabolites. The assay times were used for all the following characterization and optimization tests.
E. Further Purification of Active Metabolite:
Ammonium sulfate precipitation was performed on the active metabolites produced by strains 3A-P4, 15A-P4, and 22C-P1. This is a common method for fractioning proteins by 0 precipitation and yields a partially purified protein. The partially purified proteins obtained were utilized in further purification techniques so that a purified protein was achieved. The purified active metabolite yields a better understanding of the microbial inhibition produced by these strains.
Ammonium sulfate concentration fractions must first be determined for each strain to 5 ascertain the amount of ammonium sulfate to add to precipitate the active metabolite. Strains 3A-P4, 15A-P4, and 22C-P1 were grown separately in TSB to their respective OD.
Cell free supernatants were obtained by centrifugation at 6000 X g for 20 minutes at 4 C. Ammonium sulfate was added to the supernatant in 10% increments until a concentration of 70% is obtained.
After the addition of one of the ammonium sulfate concentrations, the supernatant was kept at 4 !O C for 2-24 h. The supernatant was harvested by centrifugation at 6000 x g for 20 minutes. The supernatant (10m1) was placed in an Amicon 10,000 MWC centrifugal device and centrifuged at 2000 x g until one ml was left in filter. This fraction was then filtered through a 0.2um filter and tested for activity against E. coli using the spot plate method. The pellet obtained from the above centrifugation process was resuspended with 10m1 of Tris-HC1 and dialyzed overnight with stirring against 2 liters of the same buffer using a Spectra/Por no. 3 dialysis tubing. It should be noted that instead of dialysis, the pellet sample can also be placed inn Amicon filter and centrifuged at 2000 x g until no liquid remains in the filter. Tris-HC1(0.05M)(10m1) is added to the filter apparatus and centrifuged at 2000 x g until no liquid remains in filter. This step is repeated and the filter is centrifuged at 2000 x g until one ml is left in filter. The preparation 4;1 was then filtered through a 0.2um filter and tested for activity against E. coil using the spot plate method. Ammonium sulfate was added to the remainder of the supernatant and the precipitation procedure repeated until the concentration of ammonium sulfate reached 70%.
The ammonium sulfate concentration that precipitates the active metabolite into the pellet after centrifugation is the ammonium sulfate percentage used to partially purify the active metabolite from each strain. Ammonium sulfate concentration needed to precipitate the active metabolite of 15A-P4 is 30%. To decrease unwanted protein and obtain a more partially purified protein it is best to add ammonium sulfate first to the sample at a lower concentration, such as 10%. This precipitates the unwanted protein, leaving the active metabolite in solution. The 30%
ammonium sulfate can then be added to precipitate the active metabolite. The supernatant is then collected by centrifugation, as previously described, and the ammonium sulfate concentration needed to precipitate the active metabolite is added.
F. Characterization of the purified active metabolite produced by 15A-P4 by gel electrophoresis:
Strain 15A-P4 was grown to its optimal OD as previously described. The crude form of the active metabolite was obtained as previously described. The active metabolite was then partially purified, after 0.1mM PMSF and 1.0M DTT were added to increase protein stability, by performing a 10% and 30% ammonium sulfate precipitation as previously described. The pellet and supernatant fractions from both percentage precipitations were kept and spot plated onto an E. coil indicator plate as previously described.
The ammonium sulfate fractions were examined using polyacrylamide gel electrophoresis (PAGE) in the presence of 0.1% sodium dodecyl sulfate (SDS) in a Mini¨Protean 3 Cell (Bio-Rad, Hercules, CA). The samples were prepared following the protocol provided with the SDS-PAGE molecular weight standards kit (Bio-Rad). A precast 10%
polyacrylamide Tris HC1 Ready Gel (Bio-Rad) was used. The current was run at 10mA constant current until the bromphenol blue entered the separating gel. Thep the current was increased to 15mA. The gel was stained using GelCode Blue stain reagent (Pierce, available from Fisher Scientific, Hampton, NH) according to the manufacture's directions. A broad range and low range standard (Bio-Rad) was used that included the following proteins and molecular weights:
Myosin (200,000), 13-galactosidase (116,250), Phosphorylase b (97,400), Serum albumin (66,200), Ovalbumin (45,000), Carbonic anhydrase (31,000), Trypsin inhibitor (21,500), Lysozyme (14,400), and Aprotinin (6,500). The 10% pellet fraction and the 30%
supernatant fraction did not inhibit E. coil during the spot inhibition assay. The 10% supernatant and 30% pellet did inhibit E. coil during the spot inhibition assay. The pellet and supernatant fractions yielded a band with a molecular weight range between 31,000 and 45,000. Therefore, this band is believed to be the inhibitory protein. (FIG. 11) G. Characterization of the active metabolite produced by 15A-P4 using low pressure column chromatography:
After ammonium sulfate precipitation, the active metabolite in the 30% pellet fraction 0 was applied to different chemistry columns to determine protein characteristics. The Bio-Rad High Q anion exchange lml cartridge and the Bio-Rad HIC
hydrophobic/hydrophilic lml cartridge were explored. To determine characteristics of the protein, 0.5m1 of the active metabolite, after ammonium sulfate precipitation was performed, was mixed with the elution buffer, and was placed on the column. A high salt buffer and low salt buffer were applied to the [5 column to determine under what conditions the active metabolite would adhere to the column.
Tris HC150mM with 10mM NaCl was used as the high salt buffer for the High Q
column and Tris HC150mM with 1.0 mM NaC1 added was used as low salt buffer for the High Q
column.
100mM sodium phosphate with no salt added was used as the low salt buffer for the HIC column and 100mM sodium phosphate with 2.4M ammonium sulfate added was used as the high salt 20 buffer for the HIC column. A flow rate of 0.7m1/min for 25 minutes was used with each buffer.
Two large fractions were collected first, one fraction was what came off the column after running a high salt buffer through the column and the other fraction was collected after a buffer containing no salt was ran through the column. These fractions were then concentrated using the Amicon centrifugal device by placing the fractions in a 10,000 MWC Amicon centrifugal device 25 and spun at 3000 rpm until dry. Two buffer washes were performed, and the protein was reconstituted to 300 1.
With the HIC column, the fraction collected during the high salt buffer application yielded a positive inhibition with the spot plate assay. The separation principle behind the HIC
column is not yet frilly understood. All theories support that interaction is related to the 30 hydrophobic surface area found on all proteins and that it is increased by high ionic strength and high temperature (11). Therefore, the fact that the protein eluted with a high concentration of salt leads to suspect that the protein is hydrophilic in nature.
With the High Q anion exchange column, both the high salt buffer fraction and the no salt buffer fraction yielded no inhibition on the spot plate assay. The procedure was repeated and nine fractions were collected after a high salt buffer was applied and nine fractions were collected after a no salt buffer was applied. Three of the fractions collected with the salt buffer showed inhibition on the spot plate assay. These three fractions also had 60 ->100mg/d1 of protein using the protein determination test. None of the fractions from the no salt buffer showed inhibition, but two fractions had 20-30mg/d1 of protein. Therefore, the fact that the 0 protein eluted from an anion column with a high concentration of salt demonstrates that the protein is a cation.
In summary, it was discovered that the active metabolite produced by strain 15A-P4 has a molecular weight between 31,000 and 45,000, is a cation, and appears to be hydrophilic.
[5 EXAMPLE 10 Determination of Stability of the active metabolites produced by 3A-P4, 15A-P4, and 22C-P1:
The stability of the active metabolite of the Bacillus isolates was assessed.
This information also helps to characterize the active metabolites. Assays were performed using the ).,0 crude form of the active metabolite to determine enzyme degradation, heat stability, and mode of action. The activity of the active metabolite was determined after exposure to enzymes and heat.
A. Enzyme degradation assay:
Enzyme degradation trials were performed on strains 3A-P4, 15A-P4, and 22C-P1 to determine ifothe active metabolite formed was a protein. Producers were grown in TSB as previously described, and the active metabolite was obtained in the crude purified form.
Enzymes (all obtained from Sigma, St. Louis, MO) used were: 6.-chymotrypsin, pronase E, proteinase K, pepsin, trypsin, and catalase. Two hundred and fifty milligrams of each enzyme 30 was added to 100m1 of sterile cold distilled water and kept on ice before use. lml of each enzyme was separately added to 4mls of the crude purified active metabolite for a final concentration of 500 pg/ml. After incubated at 37 C for 60 minutes, each sample was assayed for bacteriocin activity using the broth activity assay method. Samples without enzymes were used as controls.
The enzyme treated active metabolite was added to a 10m1 TSB tube at 10%.
Strain E.23 was used as the indicator organism and was grown as previously described. The indicator was then added to the TSB tubes, which contained the active metabolite, at a 1%
concentration.
Samples without enzyme-treated active metabolites and samples without active metabolites were used as controls. Percent inhibition was determined as previously described.
The active metabolite produced by strain 3A-P4 was found to be sensitive to a-0 chymotrypsin, pepsin, catalase, and pronase E but not affected by trypsin or proteinase K.
Active metabolite produced by strain 15A-P4 was found to be sensitive to catalase and pronase E
but not affected by 6-chymotrypsin, pepsin, trypsin, or proteinase K. The active metabolite produced by strain 22C-P1 was found to be sensitive to trypsin and pronase E
but not affected by at-chymotrypsin, pepsin, catalase, or proteinase K.

B. Heat Assay:
The temperature sensitivity of the crude purified active metabolites produced by strains 3A-P4, 15A-P4, and 22C-P1 were examined. The isolates were grown as previously indicated and the crude active metabolite was obtained. The active metabolites were heated to 100 C for ?.0 1, 5, 10, and 15 minutes; cooled to room temperature; and inoculated at 10% into 10 ml of TSB.
The active metabolites were also autoclaved for 20 minutes at 121 C and assayed for inhibitory activity. Strain E.23 was used as the indicator organism and was grown as previously described and inoculated at 1% into the 10 ml of TSB containing the active metabolite.
The assay was incubated at 37 C and OD read at either four or six hours. Percent inhibition was calculated as ?.5 previously described.
All three of the active metabolites' activity was reduced after heat treatment (Table 8).
But all three active metabolites still had some inhibitory activity after heat treatment.

Table 8: Percent inhibition of 3A-P4, 15A-P4, and 22C-P1 after heat treatments.

% Inhibition % Inhibition %
Inhibition Heat Treated No Heat Heat Treated No Heat Heat Treated No Heat 1 Min 100 C 100 100 94.4 100 88.3 Min. 100 C 75 100 87.6 100 56.3 Min. 100 C 50 100 75.3 100 46.2 100 Min. 100 C 50 100 62.0 100 25.7 100 Autoclaved 32.4 100 36.7 100 20.5 5 C. Mode of Action:
Tests were performed to determine if the active metabolites produced by strains 3A-P4, 15A-P4, and 22C-P I were bactericidal or bacteriostatic to E. coil. Strains 3A-P4, 15A-P4, and 22C-PI were grown to their optimal OD as previously described. The crude form of the active metabolite was obtained as previously described. E. coil strains E.20 and E.23 were grown as 10 previously described.
The active metabolite of each strain was tested separately and added at 10% to a 10 ml TSB tube. Strains E.20 and E.23 were added separately at 1% to the TSB tube containing the active metabolite. A TSB tube inoculated with only E.20 or E.23 at 1% was used as the control.
The tubes were incubated at 37 C, and an OD (600nm) was obtained every two hours for a total 15 of eight hours. Time zero values were also obtained. To obtain live E.
coli counts from the inoculated TSB tubes plating was performed. Serial dilutions were made every two hours for a total of eight hours and plated on TSA and incubated at 37 C overnight. Time zero values were also obtained. The plates were counted after 24h of incubation.
Strain 3A-P4 active metabolite decreased E. coli counts for both E. coli strain E.20 (FIG.
.)_13 I2A) and strain E.23 (FIG. 12B) by one log and decreased the OD
values. Strain 15A-P4 active metabolite decreased E. coli counts for both E. coil strain E.20 (FIG. 13A) and strain E.23 (FIG.
13B) by three to four logs and also decreased the OD values. Strain 22C-P1 active metabolite decreased E. coli counts for both E. coil strain E.20 (FIG. 14A) and strain E.23 (FIG. 14B) by half a log to one log and also decreased the OD values. The results indicate that 15A-P4 active metabolite is bactericidal, and 3A-P4 and 22C-P1 active metabolites are at least bacteriostatic.
Strains 3A-P4 and 22C-P1 active metabolites may also prove to be bactericidal if the assay was allowed to continue longer.

A. Media Optimization:
Strains 3A-P4, 15A-P4, and 22C-P I were grown in different media to determine a media that would yield the highest cell and spore growth. The protein found in TSB
was substituted at different percentage levels with other proteins. Carbohydrates and minerals, common to 0 industry, were also included in the different media at different percentage levels.
Strains 3A-P4, 15A-P4, and 22C-P1 were inoculated into the above different media and grown at 32 C with shaking. Samples for spore yield were aseptically removed at 24 and 48h.
The sample was placed in a 63 C water bath for 35 minutes to kill all vegetative cells. Spores were enumerated by plating serial dilutions on TSA, which were incubated for 24 hours at 32 C.
[5 Cells were enumerated by plating serial dilutions on TSA of the culture at 24h and 48h, which were also incubated for 24 at 32 C. The media that yielded the highest cell and spore growth for each strain is listed below.
B. Media Yielding Highest Cell and Spore Growth:
For 3A-P4, growth media was Primagen 2%, Sucrose 5g/L, Dipotassium phosphate ?.0 2.5g/L, 0.5g/L, MgS047H20, 0.12g/L, FeSO4 7H20, 0.05g/L, MnSO4 H20, 0.004g/L, Zn S047H20, and 0.05g/L CaCI. Growth conditions were: 32 C with shaking for 48 hours to obtain a spore count of at least 1x109.
For I 5A-P4, growth media was peptonized milk protein 5%, Dextrose 2.5g/L, Dipotassium phosphate 2.5g/L, 0.5g/L Mg504 7H20, 0.12g/L FeSO4 7H20, 0.05g/L
MnSO4 )..5 H20, 0.004g/L Zn SO4 7H20, and 0.05g/L CaCI. Growth conditions were 32 C
with shaking for 48 hours to obtain a spore count of at least lx109.
For 22C-PI, growth media was Primagen, 2%, Dextrose 2.5g/L, Dipotassium phosphate 2.5g/L, 0.5g/L MgSO4 7H20, 0.12g/L FeSO4 7H20, 0.05g/L MnSO4 H20, 0.004g/L Zn 7H20, and 0.05g/L CaCl. Growth conditions were 32 C with shaking for 48 hours to obtain a 30 spore count of at least 1x109.

Primagen and peptonized milk protein obtained from Quest International, Hoffman Estates, IL.

Field Trial A:
The objective of Field Trial A was to evaluate the ability of the selected Bacillus strains to reduce the incidence of E. coli disease and to improve performance in the nursery phase.
The site is located approximately 7 miles east of Pipestone, MN. It was a farrow to finish farm with an E. coli mortality of 20% without vaccine and antibiotic use. The intervention of vaccines and antibiotics had decreased the E. coli mortality to 5 ¨ 10%.
0 The farm consisted of one nursery barn with two rooms. Each room had two rows of six pens with each pen holding 25 pigs. Each room had a capacity of holding 300 pigs. Pigs remained in the nursery on an average of 28 days before being moved to the finishing facility.
The pigs were weaned at 26 days of age and were sorted by sex and assigned to one of three weight classes (light, medium and heavy). Control and treated pigs were placed in separate 5 rows to decrease the possibility of cross over contamination between treated and control pigs of the Bacillus strains fed to the treated pigs. Each row had three pens of gilts and three pens of barrows with one light, one medium, and one heavy weight group in each sex.
The E. coli vaccine was given to all weaned pigs. Several injectable antibiotics (gentamicin, enrofloxacin) were given to both control and treated pigs when scouring was !O observed.
The treated pigs received Product 1 in a basemix form, containing Bacillus strains and carriers as follows: 30% 3A-P4, 60% 15A-P4, 10% 22C-P1 at a final product count of 3.0 x 107 cfu/g and 40% rice hulls, 19% dried brewers grain, 40% limestone, and 1%
baylith. The basemix was then added to the standard farm pellet diet and grind and mix diet at the rate of Sibs/ton of feed to make the final Bacillus inclusion rate 7.35 x 104 cfu/g .
The pellet diet was started on day one post-weaning, and Product 1 was continued in all diet phases until the end of the nursery stage. The control pigs received the same pelleted and grind and mix diets as the treated pigs except they were devoid of the Bacillus strains.
Included in the nursery pellet diet for control and treated pigs was the antibiotic ASP-K) 250. All the grind and mix rations for both control and treated pigs included BMD and 3-Nitro.
Normal protocol was utilized for pig management.

Mortality and disease incidence was recorded weekly in both the treated and control pigs.
Pen weight, pen sex, and number of pigs in pen were recorded on day one and day 28 of the field trial.
Before the field trial began environmental, rectal, and fecal swabs were obtained from the nursery. E. coil strains from the swabs were grown and isolated at Agtech Products and kept frozen for future use. Multiplex PCR was used to determine if a strain was pathogenic.
All pathogenic E. coil isolates were individually tested in vitro against the three Bacillus strains included in Product I. This was done using broth activity assay. Each active metabolite produced by the Bacillus in Product 1 was tested against each pathogenic E.
coli strain found on 0 the farm using the broth activity assay to obtain percent degree of inhibition. The degree of inhibition was monitored by way of optical density readings using a spectrophotometer. Results of the degree of inhibition are shown below in Table 12.
Feeding Product 1 to nursery pigs during Field Trial A decreased mortality.
Mortality in the control pigs remained high at 7.0%, however, mortality in the pigs fed Product 1 decreased to 5 1.4%. Pigs had typical symptoms of E. coli disease, as seen previously and diagnosed on this farm. Therefore, the cause of mortality was determined to be attributed to E.
coli disease. No improvement in performance was seen during this trial. Treated pigs gained an average of 17.45 lbs per pig, and control pigs gained an average of 18.98 lbs per pig (Tables 9 and 10).
0 Table 9. Effect of Product 1 during Field Trial A.
Treated Control Group Group Pig Number In 139 143 PigNumber Dead 2 10 Percent Mortality 1.4 7.0 Table 10. Results of Feed Trial A.
Pen Treated Gilt or Weaning Date Pig Pen Date Pig Pen #
Weight No. Or Barrow age in # Weight Out # Weight dead Gained/pig 5/21/02 23 584 6/18/02 23 1096 22.3 5/21/02 24 479 6/18/02 24 877 16.6 5/21/02 24 425 6/18/02 23 816 1 17.8 5/21/02 23 394 6/18/02 23 729 14.6 326 6/18/02 22 729 18.3 5/21/02 23 256 6/18/02 22 573 1 14.9 5/21/02 24 290 6/18/02 24 574 11.8 5/21/02 24 372 6/18/02 23 803 1 19.4 5/21/02 25 425 6/18/02 24 869 1 19.2 404 6/18/02 22 805 1 19.0 5/21/02 24 497 6/18/02 22 902 2 20.3 5/21/02 23 540 6/18/02 18 859 5 24.2 One hundred swabs were collected from the nursery. From the 100 swabs, 100 E.
coil 5 isolates were tested to determine their pathogenicity using the multiplex PCR procedure. Fifty-three of the 100 isolates were found to contain one or more genes associated with pathogenicity.
The genotypes and the results of the inhibition of the E. coil isolates by the Bacillus in Product 1 are shown below in Table 11. All the E. coli isolates were inhibited by all three Bacillus strains ranging from 18.2 to 96% inhibition of growth. Strain 15A-P4 demonstrated the most effective 10 inhibition against pathogenic E. coil isolated from Field Trial A.

Table 11. Characterization of pathogenic E. coli isolates from Field Trial A.
E. coli Source of Multiplex Results Sample sample Bacillus Strain Bacillus Strain Bacillus Strain E. 271 Fecal STb 60.0 91.5 55.0 E. 273 Fecal STa 45.2 82.7 64.0 E. 274 Fecal STb 44.2 84.2 60.0 E. 276 Fecal K88 46.2 95.0 52.0 . .
E. 278 Fecal F18. STX2e. STa. STb 41.7 87.1 35.0 E. 279 Fecal F18. STX2e. STa. STb 36.8 86.5 37.5 E. 284 Fecal STb 44.0 85.8 53.8 E. 285 Fecal K88 38.0 86.4 51.9 E. 294 Fecal F18. STX2e. STa. STb 42.1 92.8 22.5 E. 311 Fecal STb 42.0 86.7 51.9 E. 315 Fecal STb 39.6 86.9 47.9 E. 317 Fecal STb 48.0 78.8 53.6 E. 318 Fecal K88 50.0 88.8 50.0 E. 319 Fecal F18. STX2e. STa. STb 36.1 88.8 36.8 E. 320 Fecal K88 45.8 88.8 50.0 E. 323 Fecal STb 47.9 90.8 57.4 E. 324 Rectal K88 58.3 96.8 55.8 E. 325 Rectal STb 51.9 88.8 61.4 E. 327 Rectal K99. STa 65.0 , 92.7 31.8 E. 328 Rectal K99. STa 55.0 90.4 38.5 E. 329 Rectal K99. STa 59.1 90.9 = 41.7 E. 337 Rectal F I 8. STX2e. STa. STb 41.2 90.0 23.7 E. 361 Rectal F18. STX2e. STa. STb 36.8 92.4 25.0 Environ-ment E. 374 F18. 5TX2e. STa. STb 47.4 91.9 25.0 Environ-ment E. 378 F18. STX2e. STa. STb 37.5 89.5 18.2 Env iron-ment E. 379 F18, STX2e, STa, STb 31.8 87.3 21.7 . .

In Field Trial A, feeding Product 1 to nursery swine throughout the nursery period decreased mortality due to E. coli disease. Product 1 did not enhance nursery swine performance in Field Trial A. The fact that performance was not enhanced in this trial may be due to the fact that in our laboratory testing we confirmed that ASP-250, which was included in the nursery pellet diet for control and treated pigs, is bactericidal to the Bacillus strains included in Product 1. This is most likely due to the sulfamethazine portion of this antibiotic.
Penicillin and aureomycin (chlortetracycline) have been shown in laboratory testing not to have little effect on the Bacillus strains in Product 1.

Field Trial B-1 and B-2:
The objective of Field Trial B-1 and B-2 was to develop a feed additive product containing biologically-active active metabolites from Bacillus capable of enhancing the performance of swine by reducing intestinal pathogens such as E. coli.
The site for Field Trial B-1 and B-2 was a farrow to finish facility located approximately 7 miles east of Pipestone, MN (Same site used in Example 12). A new nursery to finish facility was built in the spring of 2003. Field Trials B-1 and B-2 were performed in this new facility.
No E. coli disease was evident in the new facility.
The new facility consisted of four rooms with two large pens in each room capable of housing pigs from the nursery phase through the finishing phase. For our trials, the pens in each room were divided down the middle to make four smaller pens. Each pen could hold on the average 70 pigs.
The pigs were weaned at 18-21 days of age and were sorted by sex to one of the treatment groups. The control and treated group was comprised of one pen of barrows and one pen of gilts. The pigs remained in the stpdy for 28 days.
The treated pigs received Product 2 in a basemix form containing Bacillus strains and carriers as follows: 100% 22C-P1 at a final product count of 3.0 x 108cfu/g and 40% rice hulls, 19% distilled brewers grain, 40% limestone, and 1% baylith. The basemix was then added to the standard farm pellet diet and grind and mix diet at the rate of Sibs/ton of feed to make the final Bacillus inclusion rate 7.35 x 105CFU/g of feed. The pellet diet was started on day one post-weaning, and Product 2 was continued in all diet phases until the end of the nursery stage. The control pigs received the same pelleted and grind and mix diets as the treated pigs except they were devoid of the Bacillus strain.
The pellet diet was devoid of antibiotics. And all the grind and mix rations for both control and treated pigs included BMD, 3-Nitro, and CTC. Normal protocol was utilized for pig management.
Mortality and disease incidence was recorded weekly in both the treated and control pigs.
Pen weight, pen sex, and number of pigs in pen were recorded on day one and day 33 (Field Trial B-1) and day 31 (Field Trial B-2) of the field trial.
Feeding Product 2 to nursery pigs during Field Trials B-1 and B-2 increased performance. In Field Trial B-1 weight gained per pig and average daily gain (ADG) was 11.3%
higher in pigs fed Product 2. In Field Trial B-2 weight gained per pig was 3.9% higher in pigs fed Product 2, and ADG was 4.0% higher in pigs fed Product 2. The overall effect is summarized in Table 12. E. coli disease did not occur during these trials; therefore mortality due to E. coli was not analyzed.
Table 12. Overall effect of Product 2 for Field Trials B-1 and B-2.
Treated Control Group Group Difference Weight gained/
pig 23.48 21.8 7.71 ADG 0.737 0.683 7.91 In Field Trials B-1 and B-2, feeding Product 2 to nursery swine throughout the nursery period increased ADG by 7.9% and weight gained per pig by 7.7%. Product 2 was effective at enhancing nursery swine performance in Field Trials B-1 and B-2.

Field Trial C
The objective of Field Trial C was to evaluate the ability of the selected Bacillus strains ,to reduce the incidence of E. coli disease and to improve performance in the nursery phase.
The field trial site is located approximately 7 miles east of Pipestone, MN.
It was a nursery and finish farm with an E. coli mortality of 15% without vaccine and antibiotic use. The intervention of vaccines and antibiotics had decreased the E. coli mortality to 3-5%.
The farm consisted of two nursery barns with two rooms in each barn. Each room had four rows of six pens with each pen holding 25 pigs. Each room had a capacity of holding 600 pigs. Pigs remained in the nursery for 7-8 weeks before being moved to the finishing facility.
Pigs were placed in the nursery at 16 ¨18 days of age upon arriving at the farm and were sorted by sex and assigned to one of two weight groups (light and heavy pigs).
Control pigs were placed in one room and treated pigs in the other room to minimize the chance for Bacillus cross contamination. The E. coli vaccine was given to control pigs only.
The treated pigs received Product 1 described in Example 12 in both the standard farm pellet diet and in the grind and mix diet at 7.35 x 104 cfu/g inclusion rate.
The pellet diet was started upon placement, and Product 1 was continued in all diet phases until the end of the nursery stage. The control pigs received the same pelleted and grind and mix diets as the treated pigs except they were devoid of the Bacillus strains. None of the diets included antibiotics aimed at treating E. coli disease. Normal protocol was utilized for pig management.
Mortality and disease incidence was recorded weekly in both the treated and control pigs.
Room weight was recorded upon placement of pigs and at the end of the field trial.
Before the field trial began environmental, rectal, and fecal swabs were obtained from the nursery. Strains from the swabs were grown and isolated at Agtech Products and kept frozen for futute use. Multiplex PCR was used to determine if a strain was pathogenic.
All pathogenic E. coli isolates were individually tested in vitro against the three Bacillus strains included in Product I. This was done using broth activity assay. Each active metabolite produced by the Bacillus in Product 1 was was tested against each pathogenic E. coli strain found on the farm using the broth activity assay to obtain percent degree of inhibition. The degree of inhibition was monitored by way of optical density readings using a spectrophotometer.
Twenty days into the field trial challenges from pathogenic E. coli resulted in a death loss of 0.50% - 0.75% in the pigs fed Product 1 compared to a death loss of 3.0% -5.0% for pigs fed the control diet (no Product 1). Shortly after this period a S. suis infection became a major challenge at this farm and subsequent deaths were diagnosed at necropsy as S.
suis.
One hundred swabs were collected from the nursery. From the 100 swabs, 100 E.
coli isolates were tested using multiplex PCR procedure to identify pathogenic strains. Thirty-one of the 100 isolates were found to be pathogenic. Genotypes for each of the 31 isolates and the results of the inhibition of the E. coli by the Bacillus in Product 1 are shown in Table 13. All the E. coli isolates were inhibited by all three Bacillus strains ranging from 6.9 to 96% inhibition of growth. Strain 15A-P4 demonstrated the most effective inhibition against pathogenic E. coli isolated from Field Trial C.

WO 2005/112658 . PCT/US2005/017141 Table 13. Characterization of the pathogenic E. coli isolates from Field Trial C.
Bacillus Strain Bacillus Strain Bacillus Strain E. coli Sample Multiplex Results 3A-P4 15A-P4 22C-P1 E. 54 Rectal F18 83.5 98.5 19.2 5 E. 55 Rectal K88 80.7 95.7 14.3 E. 57 Rectal STb 97.7 99.2 52.1 E. 66 Fecal K88 69.0 91.0 19.2 E. 67 Fecal . K88 58.6 96.3 16.7 E. 69 Fecal K88 60.0 80.5 NA
E. 74 Environ- K88 53.8 95.5 6.9 ment E. 86 Rectal F18 91.0 95.8 NA
E. 87 Fecal STa, STb, K88, STx2e 66.0 NA NA
E. 90 Fecal F18 64.6 97.9 20.7 E. 91 Fecal F18 75.4 97.9 15.5 E. 96 Fecal K88 79.2 89.1 NA
E. 104 Rectal STb 92.2 95.2 37.0 E. 106 Rectal F18 89.5 96.4 NA
E. 110 Fecal Sta 79.1 79.2 NA
E. 115 Fecal STa, STb, F18, STx2e 58.9 85.3 15.0 E. 116 Fecal STa, STb, F18, STx2e 50.0 85.0 22.0 E. 117 Fecal STb 57.7 93.0 16.7 20 E. 118 Fecal K88 65.4 95.8 16.7 E. 123 Environ- STa, STb, F18, STx2e 37.0 63.1 24.0 ment E. 239 Environ- K88 93.9 95.8 15.0 ment E. 240 Fecal Sta 56.5 99.0 33.0 E. 241 Rectal F18, STx2e, STb 83.3 98.2 NA
E. 246 Fecal K88 53.8 94.2 35.7 25 E. 247 Fecal Sta 33.3 98.9 28.6 E. 251 Fecal K88, STa 39.3 90.7 23.5 E. 252 Fecal F18 97.7 99.2 30.0 =
E. 256 Rectal K88 59.2 92.8 30.0 E.257 Environ- K88, STx2e, Sib 66.3 98.5 28.6 ment E. 265 Fecal K88 81.4 96.7 12.5 E. 268 Fecal F18 96.3 99.2 25.0 In Field Trial C, feeding Product 1 to nursery swine throughout the nursery period decreased mortality due to E. coli disease. During Field Trial C, Product 1 did not enhance nursery swine performance. The lack of an improvement in performance may have been due to disease issues caused by other microorganisms, such as S. suis, and management issues.

Field Trial D:
The objective of Field Trial D was to evaluate the ability of the selected Bacillus strains to reduce the incidence of E. coli disease in the nursery phase. The site is located in Indiana. It is a 2000 sow farrow to finish farm. E. coil had been diagnosed previously by the veterinarian.
The farm has multiple nurseries. The study was performed at the Wendell Cates nursery.
The rooms consisted of two rows of 12 pens with approximately 20 pigs per pen.
Pigs remained in the nursery on an average of 35 days before being moved to the finishing facility.
The pigs came into the nursery between 11 and 13 pounds and were sorted by weight into three groups - light, medium, and heavy. Control pigs (403 head) were placed in one room and treated pigs (440 head) in another room to minimize the chance for Bacillus cross contamination.
Treated and control pigs received penicillin in the water for coughing. The control pigs received gentamycin in the water for scours.
The treated pigs received Product 3 in a basemix form containing Bacillus strains and carriers as follows: 10% of the total count of strain 15A-P4, 90% of the total count of strain 22C-P1 at a final product count of 3.0 x 108cfu/g and 40% rice hulls, 19%
dried brewers grain, 40% limestone, and 1% baylith. The basemix was then added to the standard farm pellet diet and grind and mix diet at the rate of Sibs/ton of feed to make the final Bacillus inclusion rate 7.35 x 105 cfu/g of feed. Product 3 was added to both the standard farm pellet diet and in the grind and mix diet at 7.35 x 105CFU/g inclusion rate. The pellet diet was started on day one post-weaning, and Product 3 was continued in all diet phases until the end of the nursery stage.
The control pigs received the same pelleted and grind and mix diets as the treated pigs except they were devoid of the Bacillus strains. Normal protocol was utilized for pig management.

Mortality and disease incidence, and pig numbers going into the trial and into the finisher were recorded.
Rectal and fecal swabs were obtained from the nursery during an E. coli outbreak. Three swabs were from treated pigs and three swabs came from control pigs. E. coli strains were grown and isolated at Agtech Products and kept frozen for future use.
Multiplex PCR was used to determine if a strain was pathogenic.
Treated pigs had a total death loss of 4.1%. Death loss due to E. coli disease/scours was 1.4%. Control pigs had a total death loss of 12.2%. Death loss due to E. coli disease/scours was 9.4%. Treated pigs had a placement rate of 94.5% of animals into the finisher phase compared to control pigs which had a finisher placement of 87.9% (Table 14).
Three swabs were sent from treated pigs and three swabs were sent from control pigs.
From the six swabs, 18 E. coli isolates were tested to determine their pathogenicity using the multiplex PCR procedure. The nine isolates that came from the three swabs obtained from the treated pigs were negative for pathogenic E. coli. Four of the nine isolates that came from the three swabs obtained from the control pigs were positive for the K88 pili gene and the Lt and Stb enterotoxin gene, as shown in Table 15. Therefore, the isolates from the control pigs were pathogenic E. coli and represented two of the three swabs taken from control pigs.

Table 14. Death causes during Field Trial D. Numbers represent the number of pigs that died due to that cause.
Treated Pigs Control Pigs (received Product 3) (did not receive Product 3) Scour 6 38 Small 1 1 Very Small 3 10 Fighting 1 0 Flu/Respiratory 2 0 Not Eating 1 0 Gaunt 1 0 Not Sure 3 0 Total Death Loss 18 49 Total E. co/i/Scour 6 38 Loss Table 15. Characterization of E. coli isolates obtained from Field Trial D.
Treated or Control Pigs Multiplex Results on isolates obtained from Swab swab 1 Treated Negative for E. coil virulence factors 2 Treated Negative for E.coli virulence factors 3 Treated Negative for E. coli virulence factors 4 Control Negative for E.coli virulence factors Control K88, LT, Stb 6 Control K88, LT, Stb In Field Trial D, feeding Product 3 to nursery swine throughout the nursery period decreased mortality due to E. coli disease and increased the number of pigs placed in finisher.

Field Trial E:
The objective of Field Trial E was to evaluate the ability of the selected Bacillus strains to reduce the incidence of E. coli disease and to improve performance in the nursery phase.
The study site is located approximately 7 miles east of Pipestone, MN. The cooperating producer's facility consists of approximately 450-500 sows of Babcock genetics and is d farrow to finish operation. Gilts typically enter farrowing at approximately 11 months of age and are taken through 5 farrowings. E. coli (F18) has been diagnosed on the farm with outbreaks occurring recently.
The nursery facility consists of six rooms with each room divided into one row of six pens. Each pen typically houses 25-35 pigs and pigs remain in the nursery on an average of 30 days before being moved to the finishing facility.
The pigs were weaned at 15 days of age and were sorted into one of three weight classes (light, medium, and heavy). Gilts and barrows were commingled within the same weight group in each of the pens. Ideally, the same weight group between the treated and control pens did not differ by more than 0.51bs.
A barrier was placed in between the two middle pens to divide the row into three pens of control and three pens of treated pigs. This barrier also minimized cross contamination between treated and control pigs.
The treated pigs received Product 3 as described in Example 15 in both the standard farm pellet diet and in the grind and mix diet at 7.35 x 105 cfu/g inclusion rate.
The pellet diet was started on day one post-weaning, and Product 3 was continued in all diet phases until the end of the nursery stage. The control pigs received the same pelleted and grind and mix diets as the treated pigs except they were devoid of the Product 3 Bacillus strains. The control diet did contain another commercial Bacillus product in the pelleted rations and the first grind and mix ration. Egg immunoglobins were also included in the control pelleted rations.
Included in the nursery pellet diet for control and treated pigs was the antibiotic TI 35C400 (Denagard and chlortetracycline). Normal protocol was utilized for pig management.
Mortality and other clinical signs of disease were recorded in both the treated and control pens. Comments on cause of death were also recorded. Pigs were weighed by pen using the Transcell Technology TI-500SS B. Weights were collected at weaning (day 0), day 7, and day 28. The amount of feed fed was recorded daily, and on the last day of trial all the left over feed was weighed.
Rectal and fecal swabs were obtained from the nursery during an E. coli outbreak. E. coli strains were grown and isolated at Agtech Products and kept frozen for future use. Multiplex PCR was used to determine if a strain was pathogenic.
Data were analyzed using the PROC MIXED procedure of the SAS computer program, and the effects of block and treatment, with day included, to take into account repeated measures and interactions, were evaluated. Data is summarized in Table 16.
Referring to FIGS. 15A and 15B, pig weight was influenced by treatment (P <
0.01), block (P < 0.0001) and treatment x block (P <0.01) and block x day (P <0.01).
Heavy and light pigs fed the Bacillus strains had higher body weights than pigs fed the control diet at day 28 (P <
0.005 and P <0.01, respectively), as is shown in FIG. 15B.

Average daily gain for treated pigs was always higher than ADG for control pigs. For Day 0 to 7 all main effects and the rep x block (P=0.0250) interaction was significant (FIG. 16).
The treatment x block interaction is approaching significance (P=0.102). Day 7 to 28 the effect of treatment was approaching significance (P=0.125). Day 0 to 28 all main effects were significant and the effect of treatment was nearly significant (P=0.052).
Table 16. Summary of the effect of Product 3 for Field Trial E.
Day 28 Treated Control n=18 n=18 Difference % Mortality 1.5 3.5 57.1 Weight/pig (lb) 26.98 25.89 4.2 ADG (lb/day) 0-7 0.37 0.33 12.1 7-28 0.70 0.66 6.1 0-28 0.62 0.58 6.9 Feed : Gain 1.34 1.42 5.6 Gain : Feed 0.745 0.711 4.8 Feed Intake (lb) 1927.17 1826.33 5.5 Feed intake (FIG. 17) in pigs in the light weight block was higher (P<0.01) whereas intake of pigs in the other blocks was similar (Treatment x block interaction, P < .05).
The feed:gain effect of treatment was approaching significance (P=0.125).
Referring now to FIGS. 18A and 18B, feeding the Bacillus strains reduced mortality in the high (P<0.01) and medium (P < 0.01) weight blocks at day 28 (Treatment x block x day interaction, P<0.05) (FIG. 18B).
Six swabs were sent from control pigs showing symptoms of E. coli edema disease.
None of the treated pigs displayed any symptoms of E. coil disease. Therefore, no swabs from treated pigs were taken. From the six swabs 24 E. coil isolates were tested to determine their pathogenicity using the multiplex PCR procedure. All the isolates were positive for the F18 pili gene and Stx2e toxin gene, three isolates were positive for the Sta and Stb enterotoxin genes (FIG. 19). Therefore, all the isolates were positive for pathogenic E. coil, which accounted for the clinical signs manifested by the control pigs.

In Field Trial E, feeding Product 3 throughout the nursery period enhanced performance by increasing ADG, pig weight, feed intake, and feed conversion. Feeding Product 3 throughout the nursery period decreases mortality due to E. coli disease in Field Trial E.
It is understood that the various preferred embodiments are shown and described above to illustrate different possible features of the invention and the varying ways in which these features may be combined. Apart from combining the different features of the above embodiments in varying ways, other modifications are also considered to be within the scope of the invention. For example a single Bacillus isolate, as opposed to a combination of isolates, could be used to control pathogenic swine E. coli.
The invention is not intended to be limited to the preferred embodiments described above, but rather is intended to be limited only by the claims set out below.
Thus, the invention encompasses all alternative embodiments that fall literally or equivalently within the scope of the invention.

BIBLIOGRAPHY
I. Dean-Nystrom, E. A. and Bartels-Morozov, D. 2001. Edema disease: a re-emerging problem. Proceedings of the American Association of Swine Veterinarians. 223-224.
2. Blood, D. C. and Radostits, 0. M. Veterinary Medicine 7th Edition. 637-640.
3. Heiman, R. Gayman. The Veterinary Clinics of North America Food Animal Practice.
March 2000. 117-162, 4. Bertschinger, H.U. and Fairbrother, J.M. Diseases of Swine 8th edition.
431-454.
5. Gyles, C. 2001. Escherichia coli in Diseases of Weaned Pigs: Biological Aspect.
American Association of Swine Veterinarians. 29-41.
6. Francis, D. H. 2004. Post-Weaning E. coli- diagnosis, treatment, control, and its effect on subsequent growth performance. American Association of Swine Veterinarians.

499.
7. Wills, R. W. Diarrhea In Growing-Finishing Swine. The Veterinary Clinics of North America. March 2000. 138-140.
8. Parrott, D., Rehberger, T. and Holt, M. 2002. Molecular typing of hemolytic Escherichia coli isolated from swine. Paper 385. International Pig Veterinary Society.
9. Marquardt, R. R., and et al. 1999. Passive protective effect of egg-yolk antibodies against enterotoxigenic Escherichia coli K88+ infection in neonatal and early-weaned piglets. FEMS Immunology and Medical Microbiology. 23. 1999. 283-288.
10. Nagy, B., Wilson, R., Whittam, T. Genetic diversity among Escherichia coli isolates carrying F18 genes from pigs with porcine postweaning diarrhea and edema disease.
Journal of Clinical Microbiology. May 1999. 1642-1645.
11. Roe, S. Protein Purification Techniques. Second edition. 172-175.

Claims (26)

Claims:

1. An isolated microorganism of the genus Bacillus selected from the group consisting of strains 3A-P4 ATCC PTA-6506, 15A-P4 ATCC PTA-6507, and 22C-P1 ATCC PTA-6508.

2. The microorganism of claim 1, wherein the microorganism is capable of at least one of the following:
(A) inhibiting E. coli disease in animals fed the microorganism when compared to animals not fed the microorganism, and (B) improving performance in animals fed the microorganism compared to animals not fed the microorganism, wherein the improvement in performance includes an improvement in at least one of average daily gain (ADG), weight, mortality, feed conversion, and feed intake.

3. The microorganism of claim 1, wherein the microorganism is strain 15A-P4 ATCC
PTA-6507.

4. A combination of microorganisms comprising at least two microorganisms of claim 1.

5. The combination of claim 4, wherein the combination comprises strains ATCC PTA-6508 and 15A-P4 ATCC PTA-6507.

6. The combination of claim 5, wherein the combination comprises 90% of the total count of strain 22C-P1 ATCC PTA-6508 and 10% of the total count of strain 15A-PTA-6507.

7. The combination of claim 6, further comprising a feed, wherein the combination has a count of 5.1 x 10 8 CFU/g of feed.

8. The combination of claim 5, wherein combination comprises 10% of the total count of strain 22C-P1 ATCC PTA-6508 and 90% of the total count of strain 15A-P4 ATCC PTA-6507.

9. The combination of claim 8, further comprising a feed, wherein the combination has a count of 1 x 10 6 CFU/g of feed.

10. The use of at least one strain of microorganism of the genus Bacillus, in the feeding of swine, the at least one strain being selected from the group consisting of strains 3A-P4 ATCC PTA-6506, 15A-P4 ATCC PTA-6507, and 22C-P1 ATCC PTA-6508.

11. The use of claim 10, wherein the microorganism is capable of at least one of the following:
(A) inhibiting E. coli disease in swine fed the microorganism when compared to animals not fed the microorganism, and (B) improving performance of swine fed the microorganism compared to swine not fed the microorganism, wherein the improvement in performance includes an improvement in at least one of average daily gain (ADG), weight, mortality, feed conversion, and feed intake.

12. The use of claim 10, wherein the microorganism is strain 15A-P4 ATCC
PTA-6507.

13. The use of claim 10, wherein the microorganism comprises at least two strains selected from the group consisting of strains 3A-P4 ATCC PTA-6506, 15A-P4 ATCC

PTA-6507, and 22C-P1 ATCC PTA-6508.

14. The use of claim 13, wherein the strains comprise strain 22C-P1 ATCC

and strain 15A-P4 ATCC PTA-6507.

15. The use of claim 14, wherein the microorganism comprise 90% of the total count of strain 22C-P1 ATCC PTA-6508 and 10% of the total count of strain 15A-P4 ATCC
PTA-6507.

16. The use of claim 14, wherein the microorganism comprise 10% of the total count of strain 22C-P1 ATCC PTA-6508 and 90% of the total count of strain 15A-P4 ATCC
PTA-6507.

17. A method of forming a direct-fed microbial, the method comprising:
(a) growing, in a liquid nutrient broth, at least one strain of microorganism of the genus Bacillus; the at least one strain being selected from the group consisting of strains 3A-P4 ATCC PTA-6506, 15A-P4 ATCC PTA-6507, and 22C-P1 ATCC PTA-6508; and (b) separating the microorganism from the liquid to form the direct-fed microbial.

18. The method of claim 17, wherein the microorganism is capable of at least one of the following (i) inhibiting E. coli disease in animals fed the microorganism when compared to animals not fed the microorganism and (ii) improving performance in animals fed the microorganism compared to animals not fed the microorganism, wherein the improvement in performance includes an improvement in at least one of average daily gain (ADG), weight, mortality, feed conversion, and feed intake.

19. The method of claim 17, wherein the microorganism is strain 15A-P4 ATCC
PTA-6507.

20. The method of claim 17, wherein the microorganism is strain 22C-P1 ATCC
PTA-6508.

21. The method of claim 19, further comprising:
growing, in a liquid broth, strain 22C-P1 ATCC PTA-6508; and combining strains 15A-P4 ATCC PTA-6507 and 22C-P1 ATCC PTA-6508.

22. An isolated Bacillus strain 3A-P4 ATCC PTA-6506.

23. An isolated Bacillus strain 15A-P4 ATCC PTA-6507.

24. An isolated Bacillus strain 22C-P1 ATCC PTA-6508.

25. A combination of microorganisms of the genus Bacillus, the combination comprising strains 22C-P1 ATCC PTA-6508 and 15A-P4 ATCC PTA-6507.

26. The use of the combination of claim 25 in the feeding of swine.